This study tests a putative mood-regulatory function of REM sleep and or dreaming in depressed males and females. Previous work in this laboratory has shown that those depressed who remain untreated are more likely to remit over a one year period if they demonstrate increased REM pressure (higher eye movement density, more frequent REM episodes, and reduced REM latency), and if they demonstrate higher proportions of negative affect dreams in the first half of the night than in the last (a 'working through' of affect within the night). Thirty-two depressed volunteers with equal numbers of males and females are tested for mood before and after each of two nights of laboratory-monitored sleep on three occasions over a five month period. REM sleep will be interrupted to retrieve dream reports every second night. This manipulation also increases REM pressure. Data will be analyzed for differences between sex groups for 1). the effects of REM pressure (between sleep through and REM interrupted nights) 2). Dream affect change (between half nights), 3). dream-like quality of reports, and 4). problem focus in dream content on overnight mood regulation and on remission of depression. Findings will contribute to an understanding of REM sleep and dream function, and the sex difference in rates of depression. This study will also have implications for the treatment of depression.